Complete daily slate
The day’s MLB matchups processed through the simulation engine.
MLB Season Ratings
The MLB Markov Engine
Baseball isn’t just a set of averages. This engine models the sequence of base-out states that make up an inning, then simulates each matchup 10,000 times.
US$30/month · Daily MLB slate · Model estimates are informational only.
What you get
MLB Season Ratings
Daily MLB slate
Win probabilities
Outright model probabilities for each matchup.
Projected runs
Expected run output for both teams.
Run-line & total matrix
Probability and model-implied-price views across alternate run lines and totals.
Get the product
MLB Season Ratings
US$30/month
Checkout and digital delivery are handled through Math & Bets on Gumroad.
Under the hood
How the engine is built.
01
24-state Markov chain
The engine represents baseball through its 24 base-out states and models how innings transition from one state to another.
02
Dynamic pitcher matrices
Starting-pitcher transition matrices are built from recent rolling performance, with contextual noise filtered where possible.
03
Bullpen handoff
The simulation switches from the starter to a team bullpen matrix for the late innings.
04
Extra-innings protocol
Modern extra innings are modelled with a runner beginning on second base in the 10th inning.
05
10,000 simulations
Each matchup is simulated repeatedly to form win, run and alternate-market probability distributions.
FAQ
What to know before you use it.
What makes this different from an ERA-only model?
The engine is state-based: it models how baserunners and outs evolve through an inning rather than relying only on aggregate averages.
How are starters and bullpens handled?
Starting pitchers and team bullpens use separate transition structures within the simulation.
How often is it produced?
The product is designed around the MLB daily schedule.
Math & Bets
From intuition to probability.
Use the model as a second opinion — then decide for yourself.